Cambridge Natural History Mammalia/Chapter XV

by Frank E. Beddard

CHAPTER XV

RODENTIA—TILLODONTIA

Order IX. RODENTIA^[1]

Small to moderately large animals, furry, sometimes with spines. Toes with nails of a claw-like character, or sometimes approaching hoofs. Usually plantigrade, and only occasionally and partly carnivorous. Canine teeth absent; incisors long and strong, growing from persistent pulps, and with enamel only or chiefly on the anterior face, producing a chisel-shaped edge; molars few (two to six), separated from the incisors by a wide diastema. Caecum (nearly always present) very large, and often complicated in structure. Brain, if not smooth, with few furrows, the hemispheres not overlapping the cerebellum. Surface of skull rather flat; orbits not separated from temporal fossae; malar bone in middle of zygomatic arch; palate very narrow, with elongated incisive foramina; articular surface for lower jaw antero-posteriorly elongated. Clavicles generally present. Testes generally abdominal. Placenta deciduate, and discoidal in form.

The Rodents are a very large assemblage of usually small, sometimes quite minute, creatures, embracing an enormous number of living generic types. They are distributed all over the world, including the Australian region, and, being small and often nocturnal, and by no means particular in their diet, have managed to thrive and multiply to a greater extent than any other group of living mammals. They are chiefly terrestrial creatures, and often burrow or live in ready-made burrows. Some, however, such as the Voles, are aquatic; others, e.g. the Squirrels, are arboreal, and there are "flying" Rodents exemplified by the genus Anomalurus. Their range of habitat is in fact as wide as that of any other Order of mammals, and wider than that of most.

Fig. 231.—Side view of skull of Cape Jumping Hare (Pedetes caffer). × ³⁄₅. AS, Alisphenoid; Ex.O, exoccipital; Fr, frontal; L, lachrymal; Ma, malar; Mx, maxilla; Na, nasal; OS, orbito-sphenoid; Pa, parietal; Per points to the large supratympanic or mastoid bulla; PMx, premaxilla; Sq, squamosal; Ty, tympanic. (From Flower's Osteology.)

The most distinct anatomical characteristic of the Rodents concerns the teeth. They are without exception entirely deprived of canines. Thus there is a long diastema between the incisors and the molars. Another peculiarity is, that in many cases the dentition is absolutely monophyodont. In such forms as the Muridae there seems to be no milk dentition at all. In that family there are only three molars; but in other types where there are four, five, or six molars, the first one, two, or three, as the case may be, have milk predecessors, and may thus be termed premolars. This has been definitely proved to be the case in the common Rabbit, which has the unusually large number of six grinding teeth in each half of the upper jaw when adult. The first three of these have milk forerunners. On the other hand the existence of four molars does not apparently always argue that the first is a premolar; for Sir W. Flower found that in Hydrochoerus,^[2] none of the teeth had any forerunners, at any rate so far as could be detected from the examination of a very young animal. The Rabbit appears to be also exceptional, in that the second incisor of the upper jaw and the incisor of the lower jaw have milk forerunners. In any case the tendency towards monophyodontism is peculiarly well-marked in this group of mammals. The incisors of Rodents are as a rule in each jaw a single pair of long and strong teeth, which grow from persistent pulps, and grow to a very great length, extending back within the jaw to near the hinder part of the skull. These teeth are reinforced in the upper jaw by a small second pair in the Lagomorpha only. The incisors are chisel-shaped, and often brown or yellow upon the outer face, as is the case also with some Insectivores. This peculiar shape, and their strength, renders them especially capable of the gnawing action which characterises the Rodents. It has been pointed out that where the incisors are wider than thick, the gnawing powers are feebly developed; and that on the contrary, where these teeth are thicker than wide, the animals are good gnawers. The incisors have often an anterior groove, or it may be grooves.

Fig. 232.—Molar teeth of Rodents. A, of Capybara (Hydrochoerus); B, of Squirrel (Sciurus); C, of Ctenodactylus. (After Tullberg.)

The cheek teeth vary in number from two (Hydromys) to six (Rabbit) on each side of the two jaws. Four is the prevailing number outside the large division of the Rat-like Rodents. They are often set at an angle to the horizontal plane of the jaw, looking outwards and obliquely to its longitudinal axis; the individual teeth too are not unfrequently bowed in form, reminding us of those of Toxodon. This of course only occurs in those genera which have hypselodont teeth. The pattern of the teeth varies much, and the different forms recall the teeth of more than one other group of mammals. They are either bunodont or lophodont. In many cases the tooth is encircled with a ridge of enamel, which is either almost simple or has a more complicated contour; such teeth are by no means unsuggestive of the Toxodonts. Some of the lophodont molars are by no means unlike those of the Proboscidea. In Sciurus vulgaris the encircling ridge is broken up into tubercles, which gives to the tooth a striking likeness to those of Ornithorhynchus. Other genera have teeth like those of many Ungulates. It has been shown by Sir J. Tomes^[3] that the minute structure of the enamel differs in different groups of Rodents.

The skull shows certain primitive characters. In the first place there is no distinction between the orbital and the temporal fossa.^[4] The sutures between the bones retain their distinctness for very long. Other characteristic features are the following:—The nasals are large, and so are the paroccipital processes. The palate in front of the molars is not distinct from the sides of the skull, its edge gradually becoming rounded off above. It is also very narrow. The premaxillae are large in relation to the great incisors. There is often a very much enlarged infra-orbital foramen through which passes a part of the masseter muscle. The jugal bone lies in the middle of the zygomatic arch, which is complete and enormously enlarged in the Spotted Cavy (Coelogenys paca). As in many Marsupials, the jugal bone sometimes extends backwards to the glenoid cavity, where the lower jaw articulates. It is usually said with an absolute want of accuracy that the cerebral hemispheres of the Rodents are smooth and without convolutions. This error has been repeated again and again in text-books. As a matter of fact the cerebral hemispheres of many forms are quite well convoluted,^[5] the degree of furrowing corresponding, as in so many groups of mammals, with the size of the animal. This at any rate is generally true, though the large Beaver with its scant convolutions is an exception. The smaller forms, such as Mus, Sciurus, Dipus, and Cricetus are quite smooth-brained. The best furrowed brain of any Rodent which has been examined is that of the huge Hydrochoerus. The Sylvian fissure is very generally not pronounced; but is particularly well-marked in Lagostomus. In all, or in most, Rodents the hemispheres are separated by an interval from the cerebellum, the optic lobes being visible between the two.

The mouth cavity of this group of mammals is divided into two chambers by a hairy ingrowth behind the incisors; this arrangement is useful for animals which use their strong incisors as gnawing and excavating tools as well as for the purposes of alimentation; for it allows of substances being gnawed away without the products of the chisel-like action being taken into the hinder cavity of the mouth. The Rodents have for the most part a simple stomach of normal form; but in a few this is complicated by a marked constriction, which divides the cardiac from the pyloric portions. The Hamster, for example, is thus characterised. In all the members of the order, with the exception of the Dormice and some allied forms, the caecum is large and often sacculated. In some forms (e.g. Arvicola, Myodes, Cuniculus) the large intestine is coiled upon itself in a spiral way—a state of affairs strongly suggestive of Ruminants.

The Rodents are primarily divisible into two great groups, the Simplicidentata and the Duplicidentata, characterised mainly by the upper incisor teeth. In the former there is but one pair of these teeth; in the latter a second smaller pair lie behind the former.

Sub-Order 1. SIMPLICIDENTATA.

Section 1. Sciuromorpha.

The Anomaluri are separated by Thomas and others from this section as an equal section, while by Tullberg they are grouped with Pedetes.

Fam. 1. Anomaluridae.—The genus Anomalurus suggests at first sight the Flying Squirrels of Asia, Pteromys. It is, however, an entirely African genus, and is to be distinguished from the Asiatic Rodents by a series of scales at the root of the tail, imbricated, keeled, and forming possibly a "climbing organ." This character serves also to distinguish the present genus from Sciuropterus. The cartilage, moreover, which supports the patagium springs from the elbow. There are four molars in each half of each jaw. The eyes and ears are large. There are five fingers and toes, but the thumb is small, though provided with a nail. The sternum has seven joints, and nine ribs reach it. The clavicle is strong. Huet, who has recently monographed the genus,^[6] allows six species. The species vary in size.

Anomalurus peli appears, according to Mr. W. H. Adams,^[7] to be a common species on the Gold Coast; it is coloured black and white, but in spite of the warning which this colour should convey, is considered by the perhaps rather omnivorous native as "the greatest delicacy." The animal is nocturnal, but affects only bright moonlight nights. Their "flying" consists of a jump from a high branch to a lower one, after which they reascend the tree to a point of vantage for another jump. They are said to feed upon nuts; but Tullberg only found the remains of leaves in the stomach.

Idiurus is a lately-described genus allied to Anomalurus. There are at any rate two species, I. zenkeri and I. macrotis. The thumb is more reduced than in Anomalurus, and the fibula, contrary to what is found in that genus and in most Sciuromorphs, is fused with the tibia below.

A third genus, very recently described and allied to both the foregoing, is Aëthurus. It is a native of the French Congo,^[8] and differs by the absence of flying membranes. It has, however, the pad of large scales. There is but one species, A. glirinus. It has a black bushy tail. The postorbital processes of the frontals are totally wanting—there are not even the traces to be seen in Anomalurus. The thumb has vanished. If we are to compare Anomalurus with the Squirrels then, thinks Mr. de Winton, the present genus is probably diurnal by reason of the want of flying membranes.

Fam. 2. Sciuridae.—The Squirrels, genus Sciurus, are world-wide in range, the Australian region and the island of Madagascar being alone excepted.

The eyes and the ears are large; the tail is of course long and bushy. The fore-feet have an inconspicuous thumb; the hind-feet have four toes. The soles of the fore-feet are naked or furry, those of the hind-feet are hairy. There are twelve or thirteen dorsal vertebrae, and in correspondence seven or six lumbars. The caudal vertebrae may be as many as twenty-five. In the skull the frontals are broad, and there are long postorbital processes. The infra-orbital foramen is, as a rule, not large, but is increased in size in a few forms. The number of separate pieces of bone in the sternum is five. The molars of the upper jaw are five, but the first is very small and soon drops out.

The Squirrels are often rather brilliantly coloured. The Chinese Sc. castaneiventris has grey fur with a rich chestnut-coloured under surface. The Malabar Squirrel, Sc. maximus, as its name implies, a large animal, has a deep reddish or chestnut-coloured fur above, which becomes yellow below. The "Common Squirrel," "the lytill squerell full of besynesse," which is the Squirrel of this country, is brownish red on the upper parts and white below. It ranges from this country as far east as Japan. Like many other Rodents the Squirrel likes animal food and will eat both eggs and young birds. "Camel's hair" brushes are made from this animal. The genus Tamias, almost exclusively North American in range, is included by Dr. Forsyth Major^[9] in this genus, which then consists of considerably over one hundred species.

The Ethiopian Ground Squirrels, genus Xerus, have a more elongated skull than Sciurus, and the postorbital processes are shorter. The feet are not hairy.

Nannosciurus forms a perfectly distinct genus of Squirrels. These "Pygmy Squirrels" differ in possessing a very elongated "face" and in the very broad frontal region. The teeth are unlike those of Sciurus in certain features, and have been especially compared by Forsyth Major to those of the Dormice. Four species of this genus are Malayan; one is West African.

The Bornean Rheithrosciurus macrotis is the only species of its genus. The genus may be distinguished by the exceedingly brachyodont molars, this feature being more marked in this genus than in all other Squirrels. It is called the "Groove-toothed Squirrel," from the "seven to ten minute parallel vertical grooves running down the front face of its incisors."^[10]

The genus Spermophilus includes a large number (forty or so) of Palaearctic and Nearctic animals known as Sousliks. The ears are small; there are cheek pouches as in Tamias. The general aspect of the animal is like that of a Marmot, and they bridge over the exceedingly narrow gap which separates the Marmots from the true Squirrels. Anatomically the skull is like that of Arctomys; the molars are five in the upper and four in the lower jaw. The caecum is relatively speaking very small; the measurements in a specimen of S. tredecimlineatus, dissected by Dr. Tullberg, were: small intestine, 580 mm.; large intestine, 170 mm.; and caecum, 27 mm. In Tamias also the caecum is not greatly developed. These animals are burrowing in habit.

The Prairie-dogs, genus Cynomys, of which the best-known species is perhaps C. ludovicianus, are very like the Squirrels, but they are not arboreal creatures; they live in burrows on the ground, as their vernacular name denotes. The genus is entirely North American, and four species have been differentiated.

The Prairie-dog or Prairie-marmot is some 10 inches to one foot in length. The tail is no more than 2 inches. The ears are very small; the thumb is fully developed and bears a claw. The measurements of the various sections of the intestine are the following:—Small intestine, 860 mm.; large intestine, 690 mm.; caecum, 75 mm. Thus the caecum is not large comparatively speaking. These animals dig burrows on grassy plains which they share with the Ground Owl (Speotyto cunicularis) and with Rattlesnakes, all three species appearing to live in perfect amity. Probably the Owls use the conveniently-constructed burrows, and the Rattlesnakes come there to look after the young of both.

Fig. 233.—Long-tailed Marmot. Arctomys caudatus. × ¹⁄₇.

Closely allied to the last are the Marmots, genus Arctomys. They differ in the rudimentary character of the thumb and in the longer tail. The eyes and ears are small. The distribution of the genus is Nearctic and Palaearctic. There are ten species of the genus. The Alpine Marmot, A. marmotta, is familiar to most persons. The animal lives high up in the Alps, and when danger threatens it gives vent to a shrill whistle. It hibernates in the winter, and as many as ten to fifteen animals may be found closely packed together in a single, carefully-lined burrow.

The only other European species is A. bobac, the Siberian Marmot, which occurs in the extreme east of Europe, and is also Asiatic. There are four North American species, including the Quebec Marmot, A. monax.

Fig. 234.—Flying Squirrel. Pteromys alborufus. × ¹⁄₅.

The genus Pteromys (of which the proper name, antedating Pteromys by five years, appears to be Petaurista) is confined to the Oriental region, where there are a dozen species or so. The limbs are united by a parachute extending to the toes, and supported anteriorly by a cartilage attached to the wrists. There are also membranes anteriorly uniting the fore-limbs to the neck, and posteriorly uniting the hind-limbs to the root of the tail and a trifle beyond. The skull and the dental formula are as in Sciurus, but the pattern of the molars, which is much complicated, seems to argue a different mode of nutrition. There are twelve pairs of ribs. The large intestine (in P. petaurista) is very nearly as long as the small, and the caecum is also "colossal"; the measurements in an individual of the species named were: small intestine, 670 mm.; caecum, 320 mm.; large intestine, 650 mm. The caecum is disposed in a spiral. The teats are three pairs, non-inguinal in position.

The size of these squirrels is 16 to 18 inches exclusive of the tail, which is longer. These animals can make an exceedingly long jump with the help of their flying membrane. Nearly eighty yards is the longest distance given for these aerial excursions. It is stated that they are able to steer themselves to some extent while in the air. As appears to be the case with so many Rodents, these animals feed largely upon beetles and other insects, besides bark, nuts, etc.

The allied genus Sciuropterus has a much wider range. It extends into the Palaearctic region and into North America, besides being found in India. There is here no membrane reaching to the tail. The palms and soles are furry. The caecum is very much shorter, and so is the large intestine. The latter, in S. volucella, is not more than one-third of the length of the small intestine. In other features there are no remarkable differences in structure, save that the mammae, always three pairs, may be inguinal.

Of the genus Eupetaurus^[11] but a single species is known, which is limited to high altitudes at Gilgit and perhaps in Thibet. Its principal difference from the other genera of Flying Squirrels is that the molars are hypselodont instead of brachyodont. The interfemoral membrane is rudimentary or wanting. The one species is E. cinereus. It is thought to live "on rocks, perhaps among precipices." Dr. Tullberg attributes the hypselodont teeth to the fact that the mosses upon which it is believed to feed may have much sand and earth intermingled, which would naturally lead to a more rapid wearing away of the teeth, and hence a need for a good supply of dental tissue to meet this destruction.

Fam. 3. Castoridae.—This, the third family of the Sciuromorpha, contains but one genus, Castor, the Beaver, with at most two species, one North American, the other European. This large Rodent has small eyes and ears, as befits an aquatic animal, and the tail is exceedingly broad and covered with scales; the transverse processes of the caudal vertebrae, in order better to support the thick tissues lying outside them, are divided in the middle of the series into two. The hind-feet are much larger than the fore-feet, and are more webbed than in any other aquatic Rodent.

In the skull the infraorbital foramen is small as in Squirrels. The postorbital process has practically vanished. The four molars stand out laterally from the jaws. The incisors, as might be surmised from the habits, are particularly strong. The stomach has near the entrance of the oesophagus a glandular patch, which seems to be like that of the Wombat (see p. 144). In both sexes the cloaca is very distinct and comparatively deep.

The two species of the genus are C. canadensis and C. fiber. The latter is of course the European species, which is now found in several of the large rivers of Europe, such as the Danube and the Rhone. But it is everywhere getting scarce, and limited to quite small and isolated colonies.

In this country it is absolutely extinct and has been since before the historic period. There is apparently no documentary evidence of its survival down to this period. But the numerous names of places which are called from this animal illustrate its former prevalence. Examples of such names are Beverley in Yorkshire, and Barbourne or Beaverbourne in Worcestershire. In Wales, however, Beavers seem to have persisted longer. But they were rare in the Principality for a hundred years or so before the Norman Conquest. The king Howel Dda, who died in 948 A.D., fixed the price of a Beaver skin at 120 pence, the skins of Stag, Wolf, and Fox being worth only 8 pence apiece. The Beaver was called by the Welsh "Llost-llyddan," which means "broad-tail." Its existence in the country is handed down in the name of Llyn-ar-afange, which means Beaver lake. The last positive record of the Beaver in Wales seems to be the statement of Giraldus Cambrensis that in 1188 the animal was still to be found in the river Teivy in Cardiganshire. In Scotland the Beaver is said to have continued down to a later date. Ireland it never reached. The remains of this animal by their abundance show the former prevalence of C. fiber in this country. It is known from the fens of Cambridgeshire, and from superficial deposits elsewhere. The Thames formerly had its Beavers, and apparently it was widely spread through the country generally.

The Beaver not only furnishes collars and cuffs for coats; it was used, as every one knows, to provide hats. But the usefulness of the animal by no means ended here in the eyes of our forefathers. The Rev. Edward Topsell observed that "for giving great ease unto the gowt the skinnes of beavers burned with drie oynions" are excellent. Castorein as a drug, if not in actual use, has quite recently been a part of the pharmacopoeia. It is derived from the anal glands common to this and other Rodents, and indeed many other mammals.

A large extinct form of Beaver is Trogontherium,^[12] found in the "Forest-bed" of Cromer. The skull is about one-fourth longer than that of Castor. It has a less inflated bulla, and slightly more pronounced postorbital processes than Castor. The third molar (fourth grinding tooth) is relatively larger than in Castor, and has a rather more folded crown. The foramen magnum is more triangular.

Fam. 4. Haplodontidae.—A separate family seems to be required for the genus Haplodon, whose characters will therefore be merged with those of its family. It is to be distinguished from most other Squirrel-like creatures by the fact that there is no postorbital process to the frontal. The molar teeth are five in the upper and four in the lower jaw. The Sewellel, H. rufus, like the other species of the genus (H. major), is found in North America west of the Rocky Mountains. It has the habit of the Prairie-marmot, and has a short tail, only moderately long ears, and five-toed feet. Tullberg is of opinion that this animal nearly represents the ancestral form of the Squirrel tribe.

Section 2. Myomorpha.

This subdivision of the Rodents contains, according to Mr. Thomas's recent estimate,^[13] no less than 102 genera. It is therefore obviously impossible to do more than refer to some of the more interesting of these. This group is again divided into the following families:—

(1) Gliridae, including the Dormice.

(2) Muridae, the Rats, Mice, Gerbilles, Australian Water-rats, Hamster.

(3) Bathyergidae, Cape Mole, etc.

(4) Spalacidae, Bamboo Rats.

(5) Geomyidae, Pouched Rats.

(6) Heteromyidae, Kangaroo Rats.

(7) Dipodidae, Jerboas.

(8) Pedetidae.

The Gliridae have no caecum, so usual in the Rodentia. It is true that all the genera have not been dissected, but it is known that in the true Dormice, as well as in the genus Platacanthomys, a caecum is absent.

Apart from these few exceptions the Mouse-like Rodents all possess a caecum, though it is often not very large. They are all smallish animals, and are modified to a great variety of habit and habitat. There are burrowing, swimming, and climbing forms. The group is universal in range, even including the Australian region, in which they are the only Rodents.

Fam. 1. Gliridae.—This family, also called Myoxidae,^[14] includes the Dormice, and is entirely an Old-World family, absent only from the Malagasy region. Its most important differential character is the total absence of the caecum and of any sharp boundary between the small and large intestine. The molars are usually four. The eyes and ears are well developed.

The genus Muscardinus includes only the Common Dormouse, M. arellanarius. This small creature, 3 inches long with a tail of 2½ inches, is, of course, a well-known inhabitant of this country. It is also found all over Europe. It is not particularly abundant in this country, and a good specimen is said to be worth half a guinea. As the specific name denotes, it lives largely on hazel nuts; but it will also suck eggs and devour insects. The animal makes a "nest" in the form of a hollow ball. Its hibernation is well known, and has also given rise to the German name ("Schläfer") of the group. It was well known to Aristotle, who gave or adopted the name Ἐλειὸς for the animal. Its winter sleep—suggestive of death—and its revivification in the spring gave the Bishop of Salamis, Epiphanius, an argument for the resurrection of man. The fur was reckoned in Pliny's time a remedy for paralysis and also for disease of the ears.

The genus Myoxus includes also but a single species, M. glis, the so-called "Fat Dormouse" of the Continent. It has no glandular swelling at the base of the oesophagus, such as occurs in the last genus and in Graphiurus. Of Graphiurus there are thirteen species, all African in range. The genus does not differ widely from the last. There is, however, a glandular region of the oesophagus. Eliomys is the last genus of typical Dormice. It is Palaearctic in range.

Platacanthomys, of a Dormouse-like form, has like other Dormice a long tail, on which the long coarse hairs are arranged in two rows on opposite sides towards the tip; it is represented by a single species, P. lasiurus, from the Malabar coast. It is arboreal in habit. The fur is mingled with flattened spines. The molars are reduced to three on each side of each jaw. This form has been bandied about between the "Mice" and the "Dormice"; but Mr. Thomas's discovery of the absence of the caecum argues strongly in favour of its correct location among the Gliridae. Typhlomys is an allied genus, also from the Oriental region. This and the last are placed in a special sub-family of the Gliridae, Platacanthomyinae, by Mr. Thomas.

Fam. 2. Muridae.—This family, that of the Rats and Mice in a wide sense, is the most extensive family of Rodents. In it Mr. Thomas includes no less than seventy-six genera. The molars are generally three. The tail is fairly long, or very long, and the soles of the feet are naked.

Sub-Fam. 1. Murinae.—The true Rats and Mice may be considered to form a sub-family, Murinae. The genus Mus, including the Rats and Mice in the limited sense of the word, contains about 130 species. They are exclusively Old World in range, being only absent from the Island of Madagascar. In the New World there are no species of the restricted genus Mus. The eyes and the ears are large; the pollex is rudimentary, and bears a nail instead of a claw. The tail is largely scaly. All the members of the genus are small animals, some quite minute. In this country there are five species^[15] of the genus, viz. the Harvest Mouse, M. minutus, which has a body only 2½ inches long with an equally long tail. It is the smallest of British quadrupeds with the exception of the Lesser Shrew. The Wood Mouse, M. sylvaticus, is about twice the size; it differs also from the last species in that it frequents barns, and is thus sometimes mistaken for the Common Mouse, from which, however, it is to be distinguished by its coloration and longer ears. The latter, M. musculus, is too familiar to need much description. A curious variety of it has occurred. This has a thickened and a folded skin like that of a Rhinoceros, and the hair has disappeared. The Black Rat, M. rattus, is like a large Mouse, and is smaller and blacker in colour than the "Hanoverian Rat." It is sometimes called the "Old English Rat," but seems nevertheless to be not a truly indigenous Rodent. It has been so defeated by competition with the Hanoverian Rat that it is now not a common species in this country.

The Hanoverian or Brown Rat, M. decumanus, is a larger and a browner animal than the last. It is very widely distributed through the globe, no doubt largely on account of the fact that it is readily transported by man. The same is the case with the Common Mouse, whose real origin must be a matter of doubt. The original home of the Brown Eat is thought by Dr. Blanford to be Mongolia. There is so far a justification for the name "Hanoverian Rat" that the animal seems to have reached this country about the year 1728. But there is no reason for calling it, as is sometimes done, the Norway Rat.

Some members of this genus, whose fur is interspersed with spines, or which are quite spiny, have been separated as a genus, Leggada, which, however, is not generally allowed.

Closely allied again is Chiruromys, which has a strongly prehensile tail, a feature which is not common among the Myomorpha, though Dendromys, a tree-frequenting form, and Mus minutus, already spoken of, show the same character. Many Mice seem to have prehensile tails, which they can curl round branches; but it is not so fully developed as in the species just named.

A number of other genera are referable to the true Mice, the sub-family Murinae of Thomas's classification. The Syrian and African Acomys has very spiny fur, so much so that "when it has its spines erected it is almost indistinguishable at the first glance from a diminutive hedgehog." The genera Cricetomys, Malacomys, Lophuromys, Saccostomus, Dasymys are restricted to the Ethiopian region. Nesokia is Oriental, reaching also the Palaearctic region. Vandeleuria, Chiropodomys, Batomys, Carpomys are Oriental, the last two being confined to the Philippines.

Another peculiar Philippine genus is Phlaeomys, of large size, and allied to it is Crateromys, originally confounded with it. Batomys granti is also confined to Luzon. Its molars are three, like those of the also restricted and Philippine Carpomys melanurus, which is an arboreal form. There is a second species, C. phaeurus.

Phlaeomys is placed, however, by Mr. Thomas in a distinct sub-family of its own, Phlaeomyinae, and is removed from the Murinae.

Hapalomys, with but one species, is Burmese. Pithecochirus is Javanese and Sumatran. Conilurus (also known as Hapalotis) is a genus containing species which are termed Jerboa Rats, on account of their mode of progression. They are desert and Australian forms. There are sixteen species.

Fig. 235.—Spiny Mouse. Acomys cahirinus. × ½.

Mastacomys, with one species, is limited to Tasmania. Uromys, with some eight species, is from Queensland, and inhabits also the Aru Islands and the Solomon Islands. The Celebesian Echiothrix, or Craurothrix as it should apparently properly be called, is another genus containing but a single species. Golunda is both Oriental and Ethiopian, one species occurring in each region. The beautiful little striped Barbary Mice, Arviacanthis (or Isomys), are African, north as well as tropical.

The genus Saccostomus resembles the Hamsters in the presence of cheek pouches. Its teeth, however, are Murine. It agrees with Steatomys in the comparatively short tail. The caecum is rather long.

Sub-Fam. 2. Hydromyinae.—The genus Hydromys,^[16] of which there are several species, the best known being H. chrysogaster, is an exclusively Australian form, and is aquatic in habit. It is a foot or so in length, and has a fairly long tail. The fore- and hind-limbs are webbed, in correspondence with its habits. The Australian Water-Rat is black, with an admixture of golden-coloured hairs dorsally and golden colour below, with a lighter median stripe. The thumb is small, and the webbing of the hands is not so marked as is that of the feet. The molars are only two in each half of each jaw. The caecum is rather small, the measurements of the alimentary canal being: small intestine, 895 mm.; large intestine, 278 mm.; caecum, 70 mm. Allied to the last is Xeromys, a genus which is also Australian, but limited to Queensland. It has been established by Mr. Thomas,^[17] who discovered that it has the same reduced formula as Hydromys. Xeromys, however, is not an aquatic animal, and has unwebbed feet.

In the Luzon highlands Mr. Whitehead has discovered, and Mr. Thomas quite recently described,^[18] a number of peculiar Rodents. Of these the genera Chrotomys, Celaenomys, and Crunomys are allied to the Australian and New Guinea Hydromys.

Chrotomys whiteheadi is unusual among Muridae, in its coloration being marked by a pale stripe down the back. The creature is the size of the Black Rat (Mus rattus). It is terrestrial not aquatic in habit, in spite of its likeness to Hydromys. The molars, however, are 3/3.

Crunomys fallax is more like Hydromys. It has, however, three molars, as in the last genus. But the skull has the flattened form characteristic of Hydromys as opposed to Mus.

Like Batomys, Celaenomys silaceus is also somewhat intermediate between Hydromys and Mus. It is described as very Shrew-like in appearance, and has a very pointed muzzle. Its habits Mr. Whitehead is "quite unable even to guess at." Like Hydromys and Xeromys this Rodent has but two molars.

Sub-Fam. 3. Rhynchomyinae.—The genus Rhynchomys, containing but one species, Rh. soricioides (of Thomas), is also, as both its generic and specific names imply, a somewhat Shrew-like form in external aspect. The skull, too, is Insectivore-like in its elongation and the lower incisors are worn to needle-like points. The two molars are excessively minute, and thus the always large gap in the jaws is greatly exaggerated. It is suggested that this Rat is an insect-eater, but nothing positive is known.

Sub-Fam. 4. Gerbillinae.—The Gerbilles form another sub-family, Gerbillinae, of the Muridae, or a family, according to some. The best-known genus is Gerbillus, including the Gerbilles proper. These animals are Old World in range, belonging to the three regions of that part of the world. There are a large number of species in the genus, over thirty. They have a Jerboa-like form, with rather long hind-limbs and a long and hairy tail. But the hind- as well as the fore-feet are five-toed. The molar teeth have no trace of tubercles, but only transverse lamellae of enamel. The incisors are orange; they are white in Dipus. Gerbillus pyramidum is 90 mm. long, with a tail of 125 mm. The ears are long, 13 mm. The tail has longer hairs at the tip.

Fig. 236.—Gerbille. Gerbillus aegyptius. × ½.

Psammomys is in some respects different. The tail is shorter than in Gerbillus; its length in an individual of 165 mm. was 130 mm. As in Gerbillus there are four pairs of teats, two pectoral and two inguinal. This genus is exclusively Palaearctic in range. Meriones has a range co-extensive with that of Gerbillus.

Pachyuromys is an Ethiopian genus with a short tail. As the generic name denotes, the tail is not only short but thick and fleshy.

Sub-Fam. 5. Otomyinae.—The allied genera, Otomys and Oreinomys, are Ethiopian. Otomys unisulcatus has a tail shorter than the body, the measurements of a female of this species being 137 mm. with a tail of 87 mm. The ear is long, whence the name; it measured in this specimen 20 mm.

Sub-Fam. 6. Dendromyinae.—The genus Deomys is an African form, consisting of only one species from the Congo region. D. ferrugineus has a reddish colour as its name implies; the soles are quite naked and the tail is long and slender. It is considerably longer than the body, measuring (minus a fragment of the tip) 172 mm., while the body is 125 mm. long. The characters of the molar teeth, which are three, are intermediate in their form between those of the true Rats and those of the Hamsters.

Dendromys is also Ethiopian in range. There are several species. D. mesomelas is a smallish creature, 60 mm. long, with a tail of 90 mm.

Steatomys is another African genus, allied to the last. Its tail, however, is only half the length of the body. The two remaining genera are Malacothrix and Limacomys. Their range is African.

Sub-Fam. 7. Lophiomyinae.—Allied to the Hamsters is the singular East African genus Lophiomys, with only one species, L. imhausi, of Milne-Edwards.^[19] The size is between that of a Rabbit and of a Guinea-pig. The stomach is curved and somewhat intestiniform. It has been termed the Crested Rat on account of the "prominent crest of stiff hair running down the back." The fingers and toes are five, and the very long tail is clad with hair longer than that upon the body generally. The pollex is rudimentary, and the hallux is opposable.

The most remarkable structural feature in this genus concerns the skull, and on account of this it has been regarded as the type of a separate family. The temporal fossa behind the eye is covered over by a complete bony plate, formed by a downgrowth of the parietal, meeting an upgrowth from the malar; this singular arrangement of the bones recalls the conditions which obtain in turtles. The whole skull, moreover, is covered with symmetrically disposed granulations, such as are found in no other mammal; it suggests rather the skull of certain fish. It is believed that the bony plate already referred to is not really a portion of the bones of which it appears to be a prolongation, but merely an ossification of fasciae in this region. The atlas is granulated like the skull; there are sixteen pairs of ribs and a feeble clavicle. The molars are three, and of a peculiar form. They have, in the case of the first three, transverse ridges, from which stand up two sharp and long tubercles. The other teeth have two ridges. The incisors are pale yellow. The shape of the teeth and the smallness of the caecum suggest that this Rodent is not so purely a vegetarian as others, and that it nourishes itself largely upon insects.

Sub-Fam. 8. Microtinae.—The Voles or Water-Rats form a distinct group of Murine animals, to which the sub-family name of Microtinae has been applied from the genus Microtus (more generally known as Arvicola), a genus which includes the Water-Rat and Field-Voles of this country. This genus has short ears, and a short and hairy tail. Its build is stouter and clumsier than that of the Rats. The genus is confined to the Palaearctic and the Nearctic regions. In this country there are three species. The best known is the Water-Vole or Water-Rat, M. amphibius, which has been seen by most people, and which frequents streams, ponds, and canals. The feet, curiously enough, are not webbed, which seems to argue the recent adoption of an aquatic life. Mr. Trevor-Battye has remarked that this animal, when swimming at leisure, uses its hind-limbs only, carrying the fore pair at the sides like a Seal. The Bank-Vole, M. glareolus, is rather a local species in this country. It is a terrestrial Vole, and burrows. The Field-Vole, M. agrestis, has become notorious on account of the "plagues," to which its immense numbers have on occasions given rise. It is the smallest species, and has a greyish-brown fur like the Water-Vole, the Bank-Vole being redder. To give an idea of the cost of the depredations of this animal, Mr. Scherren quotes^[20] a farmer who gave evidence before the Agricultural Commission to the effect that, putting the damage of one Vole at two pence, the amount of loss suffered on a farm of 6500 acres in two years would be £50,000!

The genus Fiber comes very near the last. It is a North American genus. The hind-feet are slightly webbed; the tail is a trifle shorter than the body, and is compressed and scaly, with scattered hairs. The thumb is short, but with a fully-developed claw. As in the last genus, the small and large intestines are roughly of the same length, and the caecum is about one-fourth of the length of either. It is known as the "Musquash."

Of Fiber zibethicus, or rather a closely-allied form, F. osoyoosensis, from Lake Osoyoos near the Rockies, Mr. Lord writes^[21] that it constructs for itself a house of bulrushes built up from the bottom in 3 or 4 feet of water. It is dome-shaped, and rises about a foot out of the water. "If a dead or badly-wounded duck be left on the pool, it is at once seized on, towed into the house, and doomed." Thus it appears that this Rodent, like so many others, is largely carnivorous. It has also been asserted that it eats fish.

Neofiber is an allied genus, North American in range. The species, N. alleni, is compared, as regards outward form, with the Water-Vole, M. amphibius. It has, however, a shorter tail.

Another very well-known member of this sub-family is the Lemming. The name, however, applies to two quite distinct genera. The genus Cuniculus, including the Banded Lemming, C. torquatus, is an inhabitant of North America, Siberia, and Greenland. The tail is short, its length being 12 mm. as against a body length of 101 mm. The feet are furred beneath, a not unusual state of affairs in Arctic mammals. The ears are very slight. The thumb is well developed, and bears a claw.

In Myodes, on the other hand, which is not so markedly an Arctic animal, though occurring in both Palaearctic and Nearctic regions, the ears are rather bigger, though still smaller than those of Microtus. The under surfaces of the feet are similarly furred. The tail is also short. It is commonly said that the two genera are to be distinguished by the furred feet of Cuniculus, and by the absence of fur in the present genus. That, according to Tullberg, does not appear to be the case. The differences are thus so much reduced that it seems almost unnecessary to retain the two genera. The best known species of Myodes is of course the Scandinavian Lemming, M. lemmus. This animal used to occur in this country in Pleistocene times (as did also C. torquatus), and recently Dr. Gadow has found remains with skins attached in caves in Portugal. It may still survive in the mountains of the Peninsula.

The actual habitat of the Lemming in Scandinavia is the great tablelands, 3000 feet high in the centre. The migrations do not take place with regularity; even twenty years may elapse before the appearance in cultivated lands of those countless hordes so familiar (as far as their description is concerned) to everybody. The Lemmings do not return from their exodus. They die from various causes, including combats with one another. Their chief foes, however, are Wolves and Gluttons, Buzzards and Ravens, Owls and Skuas, which batten on the migrant hordes. Their sudden increase in numbers recalls the similar increase at times of the Field-Vole, to which reference has already been made.

Ellobius is an Old-World genus, which leads a "Talpine" life, and has in consequence rudimentary external ears and very small eyes. The tail is short. Contrary to what might be expected from its mode of life, the claws upon the digits are not strong.

The remaining genera of Vole-like Murines are Phenacomys and Synaptomys from North America, and Siphneus from Palaearctic Asia. Evotomys is one of those genera which are common to both the Palaearctic and the Nearctic regions, but the bulk of the species are North American.

Sub-Fam. 9. Sigmodontinae.—This is the name given to another sub-family of Murine Rodents, a group which includes the Hamsters in the Old World as well as a large number of South American genera of Rat-like animals. Of these latter there are a very large number, the bulk of the group being American.

The Hamsters, genus Cricetus, as it is usually called, although apparently the correct name is Hamster, are Old-World forms of Pouched Rats. The Common Hamster, C. frumentarius, is about 210 mm. long, with a tail of 58 mm. It has cheek pouches. The small and the large intestines are not very unequal in length, and the caecum is fairly large, being about one-sixth to one-seventh of the length of either. It is a purely vegetable-feeding creature, and in Germany where it occurs (and from which language its vernacular name is derived), hibernates during the winter in its burrow, having previously surrounded itself with a great accumulation of food carried thither.

To North America are peculiar the genera Onychomys, Sigmodon, and Peromyscus. The genus Sigmodon, the Cotton Rats, reaches Central America, and even gets a little farther south. The other two genera, though mainly North American, also extend their range to the south. Onychomys has hairy footpads, a state of affairs which characterises a number of these Rodents.

The genera Megalomys, Chilomys, Reithrodontomys, Eligmodontia, Nectomys, Rhipidomys, Tylomys, Holochilus, Reithrodon, Phyllotis, Scapteromys, Acodon, Oxymycterus, Ichthyomys, Blarinomys, Notiomys are South American forms. Oryzomys and Rheithrodontomys are common to both parts of the New World.

The genus Ichthyomys is remarkable on account of its un-Rodent-like habits and of certain associated structural changes. I. stolzmanni was obtained from Mount Chanchamays in Peru at an altitude of 3000 feet; it is a habitual fish-eater, and lives in streams. Another species, I. hydrobates, was formerly referred to Habrothrix. The skull shows likenesses to that of the Australian Hydromys; but the most marked characters of adaptation are those of the teeth and caecum. The cutting edges of the upper incisors form a reversed of obvious use in holding a slippery fish. The caecum is much reduced, short, and narrow. The general Otter-like shape of the creature is largely due to its flattened head, though its "size and general proportions are much as in the common Black Rat."^[22]

This sub-family contains a number of genera from Madagascar, viz. Brachytarsomys, Nesomys, Hallomys, Brachyuromys, Hypogeomys, Gymnuromys, and Eliurus.

Sub-Fam. 10. Neotominae.—The last sub-family of the Muridae is that of the Neotominae, containing the North American genera Neotoma, Xenomys, Hodomys, and Nelsonia.

Fam. 3. Bathyergidae.—This family contains several genera which consist of subterranean forms. All these Rodents agree in a number of characters, of which the principal are as follows:—

The eyes are very small, and the external ears are reduced to the merest fringe of skin round the aural aperture. The legs are short, as is the tail; the hair-covering is reduced—a reduction which finds its culmination in the nearly nude Heterocephalus. Being burrowing creatures, a number of other modifications in accordance with this mode of life are to be seen in their structure. The upper incisors stand out in front of the closed lips, and prevent the entrance of earth. For the same reason the nostrils are small, and the forehead but little expanded between them.

The genus Bathyergus contains but a single species, the Cape Mole-Rat, which is found in Southern Africa; it is of moderate size, not exceeding a small Rabbit in dimensions. On the fore-limbs are exceedingly long claws, of which that borne by the second finger is the longest, and the claw of the thumb the shortest. The hind-feet have by no means such long claws. The scratching and burrowing is naturally chiefly effected by the fore-limbs. The small and large intestines are of equal length, and each is rather more than six times the length of the caecum; in these measurements the present genus differs from the next.

Georhychus.—Of this African genus there are about ten species. The claws are not so long as in the last genus, but there are, as in Bathyergus, four molar teeth on each half of each jaw. The intestinal measurements in an example of G. capensis were: small intestine, 25 inches; caecum, 4 inches; large intestine, 15 inches.

The genus Myoscalops or Heliphohius (also with an African range) has six back teeth on each side. A number of species sometimes referred to the last genus are placed here by Mr. Thomas. The claws are small.

One of the most remarkable genera of this family is the little Heterocephalus from Abyssinia and Somaliland. As Mr. Thomas justly remarks,^[23] it "is a peculiar-looking little creature, about the size of the Common Mouse, but looking almost more like a tiny hairless puppy on account of its nearly naked skin, small eyes, and peculiar physiognomy." Though apparently naked, there are numerous scattered hairs over the entire body, and the toes are fringed with stiffish hairs, which must be advantageous to a burrowing animal. There are two species, H. glaber (originally described by Rüppell), and H. phillipsii, of which our knowledge is due to Mr. Thomas. The length of the entire creature including the tail is not more than 134 mm., both species being approximately of the same dimensions. Mr. Lort Phillips, the discoverer of the species which bears his name, writes "that this little creature, called 'Farumfer' by the Somali, throws up in places groups of miniature craters, which exactly resemble volcanoes in active eruption. When the little beasts were at work, I used frequently to watch them, and found that the loose earth from their excavations was brought to the bottom of the crater, and sent with great force into the air in a succession of rapid jerks, and that they themselves never ventured forth from the shelter of the burrows."^[24]

Fam. 4. Spalacidae.—"The Spalacidae," observes Dr. Blanford, "are sometimes called rodent moles, and resemble a mole in general aspect, having cylindrical bodies, short limbs, small eyes and ears, large claws, and a short or rudimentary tail." The existence of a spiral valve in the caecum may perhaps characterise this family; but it has at present only been found in the two genera Spalax and Rhizomys.

Spalax has inconspicuous eyes and external ears. The tail is totally absent. The lower incisors are more developed than in other Rodents; they project in a bony sheath beyond the posterior end of the ramus of the lower jaw. The scapula is long and narrow. The large intestine is half the length of the small intestine. The animal seems to have only two pairs of teats, one pectoral the other inguinal.

Spalax typhlus of Egypt, which is probably not different from the European form, makes extensive burrows, some of the branches being even 30 to 40 yards in length. In a "domical chamber," situated along the course of one of these burrows, Dr. Anderson found no less than 68 bulbs stored up. Its eyes are mere black specks among the muscles, but they appear, however, to have a proper organisation. There are altogether eight species of the genus, which is entirely Palaearctic in its range.

The genus Rhizomys, including a number of species known as Bamboo Rats, is purely Oriental in range. Rh. sumatrensis reaches a length of 19 inches; the better-known species, Rh. badius, is at most only 9 inches in length—in both cases the measurements are exclusive of the tail, which is a quarter to one-third of the length of the body, and is not scaly but nearly naked, with a few scattered hairs. The molars are three, and the incisors usually orange in colour; but sometimes the upper incisors are white as in Rh. badius. There are thirteen dorsal vertebrae. In Rh. pruinosus the large intestine is considerably longer than the small intestine; the lengths of the two sections of the gut are 42 and 30 inches respectively. In another species the large intestine is slightly shorter than the small intestine. In Rh. badius the two parts of the gut are almost exactly equal in length. There are three pairs of inguinal and two pairs of pectoral teats. The name Rhizomys appears to have been given to the animals of this genus for the reason that they feed largely on roots. They burrow, and, like many other burrowing animals, feed in the evening. As is the case with other forms, Rhizomys is said to burrow with the assistance of its teeth as well as of its claws.

Fig. 237.—Bamboo Rat. Rhizomys badius. × ¼.

Tachyoryctes is an African genus closely allied to the last. There are three Ethiopian species. It is mainly to be distinguished by the different pattern upon the grinding surface of the molars.

Fam. 5. Geomyidae.—This family of burrowing Rodents is limited to North and Central America. The animals have cheek pouches, and small eyes and ears, in accordance with their mode of life. The claws of the fore-limbs are very strongly developed.

The genus Geomys contains some eight species, which are Central and North American, not extending, however, far north. The incisors of the upper jaw are grooved with two grooves. There are three pairs of teats—one axillary, and the two remaining inguinal.

Thomomys, without grooves on the incisors, reaches to Canada in the north, and does not extend as far south as the last genus.

Allied to this family, and indeed united with it by Tullberg, but kept separate by Thomas, is the

Fam. 6. Heteromyidae.—The members of this family are also American, but are not confined to the northern-central regions of that continent, for the genus Heteromys extends into South America.

The genus Dipodomys, with twelve species, is of a Jerboa-like form, as the following measurements of an example of D. merriami will show. The length of the head and body was 85 mm.; of the tail 127 mm.; the hind-foot is 32 mm. It has but four toes. The hind-limb is longer than the front-limbs.

In Perodipus the same form is exhibited. There are, however, five toes, and the sole of the foot is hairy. The axis vertebra and the two following vertebrae are fused together.

Perognathus is a third genus. It has the same general slender form, but the tail is not so long, being but little longer than the body. The hind-limbs, too, are shorter. The teats of this and of Perodipus are as in Geomys. The two remaining genera of the family are Heteromys and Microdipodops.

Fig. 238.—Jerboa. Dipus hirtipes. × ⅓. Eastern Europe.

Fam. 7. Dipodidae.—This family consists of small, plain-living, and leaping or arboreal creatures, commonly known as Jerboas. The main anatomical characters of the family are the following:—There is a large infra-orbital foramen. The molars are always reduced, the premolar being either absent in the lower jaw alone or in both jaws. This family presents an obvious likeness to Dipodomys (hence the name of the latter) and to some other members of the American family Heteromyidae There is even the same ankylosis of the neck vertebrae. We find, moreover, the same association of long-legged and shorter-legged forms that characterises the Heteromyidae.

Fig. 239.—Bones of right pes of Jerboa, Dipus aegyptius. × ¾. a, Astragalus; c, calcaneum; c², middle cuneiform; c³, outer cuneiform; cb, cuboid; n, navicular; I-IV, first to fourth toes. (From Flower's Osteology.)

The typical genus Dipus is a smallish quadruped with long naked ears and a long tail. The ten species are all Palaearctic in range. The fore-limbs are short and five fingered, and the short pollex has no claw; the hind-limbs are excessively long and only three-toed. The bony structure of these limbs is remarkable. The three metatarsals are elongated almost like those of a bird, and are ankylosed together. The digits have long phalanges which alone reach the ground as the animal hops. It is a curious fact, and one not so easily identifiable with the way of life, that the neck vertebrae of this genus are ankylosed together with the exception of the atlas, which is free; the arrangement is precisely like that of the Sperm Whale. The last vertebra is, however, sometimes free. The Jerboas not only leap but they burrow, and their strong incisors are said to be used in burrowing through stony ground. They are eaten by the Arabs, and are, or have been, called Daman Israel, i.e. Lamb of Israel. In D. hirtipes the body and tail measure respectively 4½ and 7 inches. The hind-feet have a tuft of long hairs below. Mr. W. L. Sclater's newly-founded genus Euchoreutes^[25] is somewhat more primitive in its characters than is Dipus. The general form is the same, with long ears and a long tail. But there are five toes to the hind-limb, the two lateral ones though nailed being much shorter than the middle three. It has a "long pig-like snout," and the tail is cylindrical as in most other Jerboas, with a tuft of longer hairs at the end. The incisor teeth, grooved in Dipus, are here smooth, as in Alactaga. The species was probably obtained "in the sandy plains round the city of Yarkand."

Alactaga is much like Euchoreutes; it has five toes, a cylindrical tufted tail, the hairs at the end distichous, smooth incisors, and a premolar present in the upper jaw. It also differs from Euchoreutes by the much smaller auditory bulla as well as in the fact that the infra-orbital foramen has no separate passage for the nerve, which passage is to be distinguished in both Dipus and Euchoreutes. The best-known species is the Siberian Jumping Rabbit, A. jaculus. Beneath the ends of the three main toes of the feet are remarkable fan-shaped pads. In A. decumana the body and tail measure 7 and 10 inches respectively, the ears 2 inches. Platycercomys, a fourth genus of the family, is much less known and is to be differentiated from the last three genera by the fact that it has no premolars at all, the grinding tooth formula being thus 3/3. The tail too is flattened and "lancet shaped." It extends from Siberia to Nubia, and thus just enters the Ethiopian region.

The above are the more typical Jerboas. There remain several forms which are not at all Jerboa-like in their way of life, but are nevertheless, on anatomical grounds, placed with them. Zapus, an American genus, with the exception of one Palaearctic species, is transitional in that its hind-legs are rather long, but there is not so much difference between them as in the typical Dipodidae. Sminthus is at the opposite extreme to Dipus. Its feet are short and of equal length; it climbs in trees, and may perhaps be looked upon as nearest of all Dipodidae to the ancestral form of the group.

Fam. 8. Pedetidae.—The genus Pedetes contains but one species, P. caffer, the Cape Jumping Hare. The animal suggests a large Jerboa in appearance on account of its jumping habits, the long hind-limbs, and the long tail. The length of a fair-sized example is some 17 inches, with a tail of the same length. The eyes and ears are large. The hands are five-fingered and the feet only four-toed, the hallux being of course the absent digit. In the skeleton it is interesting to note that the second and third cervical vertebrae are so close together that there can be no free movement; interesting because in Dipus the cervicals are actually ankylosed. The dorsal vertebrae are twelve. The small intestine is long, measuring 7 feet 4 inches, while the caecum is short, being only 8 inches long. The large intestine is 3 feet 10 inches long. The gall-bladder appears to be absent,^[26] an exceptional state of affairs in Rodents. A singular fact in the anatomy of this animal is the existence of a septum dividing the lower part of the trachea. This is sometimes met with in birds. As might be supposed from its large eyes, the Spring Haas, as the animal is sometimes called, is nocturnal. Its long hind-limbs permit it to leap enormous distances. It is a burrowing Rodent.

Section 3. Hystricomorpha.

Fam. 1. Octodontidae.—The Rodents of this family are of small to moderate size, the only, relatively speaking, giant in the family being the "Water-Rat," Myocastor. The toes are with but one exception not reduced; the tail is long in the majority of the genera. The teats are placed high up on the sides of the body. The clavicle is fully ossified. All the genera are South or Central American in range with the exception of Petromys.^[27]

Sub-Fam. 1. Octodontinae.—Octodon has four species, which are all Chilian, Peruvian, and Bolivian in distribution. The Degu, O. degus, has a length of 160 mm., with a tail 105 mm. long. The ears are 18 mm. long. At the roots of the claws are longish and stiff hairs which appear to serve as "combs." The tail has long but sparsely scattered hairs. There are twelve pairs of ribs. The lengths of the various sections of the intestine are as follows: small intestine, 680 mm.; caecum, 90 mm.; large intestine, 390 mm. These animals live in large companies. Closely allied is the genus Habrocoma (more correctly, as it appears, to be written Abrocoma), with two species. H. bennetti is 204 mm. long, with a tail of 103 mm. The ears are long, 22 mm. The fore-feet have no outward trace of the thumb. Stiff hairs like those that characterise Octodon are found also in this genus. The fur is very soft. The furring of the tail is much thicker than in Octodon.

Spalacopus with but a single species, S. poeppigi, is a burrowing animal, from which indeed, and on account of its resemblance to Spalax, it has received its name. The ears in accordance with the underground life are short, only 5 mm. in length in an example of 120 mm. The tail too is reduced, being in the same example only 42 mm. in length. As in the last two genera the large intestine is about one half of the length of the small intestine.

The "Tuco-tuco," genus Ctenomys, has also short ears and tail. The claws of the fore-feet are longer than those of the hind-feet.

A related form is Aconaemys (better known as Schizodon), with similar external characters; it inhabits high localities on the Andes.

Petromys is the only genus of the sub-family which is not American in habitat. It is an African form and there is but one species. Its anatomy conforms to that of the genera already considered. The main difference in structure is shown by the teeth. Their surface is uneven, and differs from that of other Hystricomorphs "in that the enamel to the inside of each upper jaw-tooth and outside on each lower jaw-tooth forms two tubercles, to which correspond grooves in the reverse position of the applied teeth."

Sub-Fam. 2. Loncherinae.—The genus Echinomys with thirteen species belongs to the Neotropical region. The members of the genus are entitled "Spiny Rats" since they have spines mixed with the fur. The tail is long and the ears are very well developed. Both feet are five-toed. The tail is scaly as well as haired. Trichomys (also called Nelomys) is very close to the above, and is also from the same part of the world.

The genus Cannabateomys contains but one species, C. amblyonyx, which was formerly included in the genus Dactylomys, but has lately been separated by Dr. Jentink.^[28] The animal is Brazilian and has a total length of 520 mm., of which 320 mm. belong to the tail. It is a climbing rat, and in accordance with that way of life has undergone some modifications. The fore-feet are four-toed, the two middle toes being markedly longer than the outer ones. The hind-feet are five-toed with the same greater development of the two middle toes. The claws are small and somewhat nail-like.

Dactylomys, also Brazilian, and with but one species, D. dactylinus, differs from the last in the fact that the molars are simpler in form; they are divided into two lobes, each of which has but a single enamel fold, whereas in Cannabateomys these teeth have several enamel folds. The tail, moreover, is but slightly hairy.

Loncheres with eighteen species is another Neotropical genus allied to the foregoing. Small spines are, as in many of these genera, intermingled with the fur. This genus has as many as seventeen dorsal vertebrae, which is an unusually large number. L. guianae is known as the "Porcupine Rat." Allied genera, also South American, and without spines in their fur, are Mesomys, Cercomys, and Carterodon.

The South American Thrinacodus is also known by one species,^[29] T. albicauda, which has rather more than the distal half of the long tail of a white colour. The fore-feet have four toes. The ears are broad and short.

Sub-Fam. 3. Capromyinae.—A third sub-family of the Octodontidae is formed by the genera Myocastor, Capromys, Plagiodontia, and Thrynomys, which are all Neotropical forms with the exception of the last, which is African.

Thrynomys (better known perhaps as Aulacodus) is a genus of African Rodent, containing some four species. The best-known of these is T. swindernianus, the Ground-Rat of West and South Africa. Its structure has been investigated by Garrod,^[30] by Tullberg,^[31] and by myself.^[32] The fur is mingled with flattish bristles; the tail is moderately long, about half as long as the body. The fore-feet are five-toed, but the two toes at each end of the series are quite small. The hind-feet are only four-toed, the hallux being absent. The claws of the hind-feet are stronger than those of the fore-feet. The ears are not long. The limbs are decidedly short, hence the name of "Ground-Pig" sometimes applied to this animal. The molars are four in number in both jaws. The incisors of the upper jaw are twice grooved. There are thirteen dorsal vertebrae. The length of the small intestine is 60½ inches, that of the large 49; the caecum is short, being only 8 inches long. It is a remarkable fact that the acromion is joined to the rest of the spine of the scapula by a joint.

Myocastor, a name which seems to have the rights of priority over the more familiar Myopotamus, applies to a large South American aquatic Rodent. The general aspect of the animal suggests a Water-Rat of large size (it has been exhibited in shows as a phenomenal product of London sewers!); the tail is nearly as long as the body. The ears are small. The limbs are short. The tail is naked. The hind-feet are webbed, but not so much so as in Hydromys. A small thumb is present. The animal has thirteen pairs of ribs; the molars are four in each jaw. The large intestine is more than three times the length of the small, and the caecum is, as in the last genus, relatively short.

Capromys is a genus^[33] which is remarkable on account of its restricted distribution. It is found only in the islands of Cuba and Jamaica. There are four species, of which C. melanurus is a dark brown-coloured animal with a blacker tail, nearly as large as a Rabbit. The native name of this Rodent is "hutia." It is also remarkable for having a stomach more complicated than is the rule among the mammals of this group. The organ is divided by two constrictions into three compartments. In C. pilorides the liver is occasionally divided up in an extraordinary fashion into small lobules. Capromys has the large number of sixteen dorsal vertebrae.

Fam. 2. Ctenodactylidae.—For these African genera it seems admissible to form a distinct family, though Thomas, and Flower and Lydekker, only allow to the genera Ctenodactylus, Pectinator, and Massoutiera sub-family rank. On the other hand, Tullberg removed these genera entirely from the Hystricomorph section and placed them as a section of the sub-tribe Myomorphi of the tribe Sciurognathi. It was chiefly the form of the mandible which led to this placing, for in these Rodents, as in all Squirrel- and Rat-like Rodents, and unlike what is found in the Hystriciform genera, the angular process of the mandible is not bent sideways.

The genus Ctenodactylus derives its name from the peculiar strong bristles which form a comb-like structure upon the hind-feet and hide the claws; these are stated to be for the purpose of dressing the fur. The Gundi of North Africa, C. gundi, has a length of 190 mm., with a short tail of 17 mm. The ears are only moderate in size. The dental formula of the molars is 4/3. The incisors are white. The feet have four digits, and the hind-limbs are the longer. The large intestine is distinctly longer than the small intestine.

Pectinator spekii is the only representative of a genus not far removed from Ctenodactylus; it is a smallish Rodent, 6 inches in length, exclusive of a rather bushy tail nearly 3 inches long. It comes from Abyssinia. It has somewhat the appearance of a Squirrel, which is heightened by the fact that when sitting the tail is arched over the back; when running the tail is carried out straight. There are only four toes visible externally on both fore- and hind-limbs, but pollex and hallux exist in the skeleton, with a single phalanx each. There is only a single pair of mammae, and in correspondence with this but two or three young are produced at a time. The hind-feet have bristles very much like those of Ctenodactylus. The molars, however, are 4/4. There are twelve ribs, of which six reach the sternum. The latter is made up of six pieces, and the manubrium in its breadth anteriorly suggests that of the Vizcachas. The clavicles are present.^[34]

Fig. 240.—Carpincho. Hydrochoerus capybara. × ¹⁄₁₂.

Fam. 3. Caviidae.—This family, which includes the Cavies and the Capybara, is entirely South American and West Indian in distribution. It embraces animals of fair to large size, the Capybara (or Carpincho) being the greatest of existing Rodents. The ears are well developed. The toes are commonly reduced, and the members of this family possess only a rudimentary tail. The hair though rough is not spiny. Other characters had best be deferred until the several genera are treated of. We shall begin with the giant of the family, the genus Hydrochoerus. This genus contains but a single species, H. capybara of South America. It reaches a length of some 4 or 5 feet. The ears are not large; the tail is completely absent. The fore-feet are four-toed, the hind-feet three-toed; the digits are webbed, though not to a very great degree, and the nails have the appearance of hoofs. There are fourteen dorsal vertebrae; the clavicle is absent. In the skull the paroccipital processes are of great length. The infra-orbital foramen is large. The most remarkable fact about the teeth is the great size of the posterior molar of the upper jaw; it has fourteen folds of enamel, more than all the anterior teeth possess collectively. The incisors are white and grooved in front. The measurements of the alimentary tract as given by Tullberg are: small intestine, 4350 mm.; caecum, 450 mm.; large intestine, 1500 mm.

Fig. 241.—Patagonian Cavy. Dolichotis patachonica. × ¹⁄₁₀.

The Capybara or Carpincho is largely aquatic in its habits. Their "favourite locality," writes Mr. Aplin,^[35] "is a broad laguna in the river, furnished with open water, and also beds of 'camelotes,'—a sloping open grassy bank on one side, where the Carpinchos can lie in the daytime in the cooler weather, sleeping and basking in the sunshine; on the other a low shelving bank, clothed with 'Sarandi' scrub growing out into the black reeking mud and shallow water beyond." They always take to the water when alarmed, at a rate and with a gait which reminded Mr. Aplin of a Pig. When in the water they swim slowly with the upper part of the head, including nose, eyes, and ears, above the surface. But they can dive for a considerable time and distance, and baffle their enemies by seeking the shelter of a mass of water-plants, and lying there with their noses only just above the surface.

The genus Dolichotis^[36] has long ears, and generally resembles a rather long-legged Hare in appearance. The front-feet are four-toed, the hind three-toed. The Patagonian Cavy, as this animal is called, has twelve dorsal vertebrae, and rudimentary clavicles.^[37] The paroccipital processes are long; the incisors are white, and are not grooved in front. The sternum has six pieces, and seven ribs reach it.

Cavia, including the species C. porcellus, the Guinea-pig (which name is a corruption apparently of Guiana pig), has the same number of toes on its hind- and fore-feet as has the Capybara. The name applied to the wild stock whence our Guinea-pig is derived is the Restless Cavy. The fur is greyish; of the domestic animals the colour is too well known to need description.

Fam. 4. Dasyproctidae.—The genus Coelogenys includes but two species. C. paca, known as the "Spotted Cavy" or "Paca," has a brown body, with white spots like those of a Dasyure; it is one of the largest of Rodents, and has a quite short tail. The hand and foot are both provided with five digits; but the thumb is small, and in the foot the three middle toes considerably exceed the others in length. The hind-foot is practically three-toed. The fibula is not nearly so reduced as in Dolichotis. The skull of the animal is remarkable for the extraordinary development in breadth of the jugal arch, which is sculptured externally. There is a large cavity formed below, at the maxillary end of this huge arch, by the curving inwards of the bone, which lodges a cavity continuous with the mouth. The palate has anteriorly a ridge on either side, and is thus divided from the sides of the face in a way which is not found^[38] in the allies of Coelogenys. Clavicles are present. There are thirteen dorsal vertebrae. The incisors are coloured red in front. The animal is South American, and in that continent is limited to the Brazilian sub-region. This, the best-known species of Paca, is called the Gualilla by the natives of Ecuador; in the same district another form is met with which the natives term Sachacui (signifying Forest Cavy). It is very often the case that a different native name expresses a real specific difference; and to the latter form M. T. Stolzmann has given the name of C. taczanowskii.^[39] This form, unlike the common Paca, which is fond of forests and low-lying ground in the neighbourhood of water, is alpine in habitat, living upon mountains of 6000 to 10,000 feet. It burrows in much the same way as its congener, and is greatly sought after as food, its meat possessing an "exquisite taste." It is pursued by dogs, by whose aid one of the two entrances to the burrow is guarded, and the creature is smoked out and killed with a stick.

The genus Dasyprocta, containing those Rodents known as Agoutis, is divisible into several species, apparently about twelve, all of which are, like the Pacas, confined to the Neotropical region. They have, however, a much wider range within that region, and occur as far north as in Central America and in some of the West Indian Islands. They are of rather smaller size than the Paca, and are without spots. The colour is of a golden brown in some forms, but usually has a freckled, grizzled, greenish kind of appearance. The tail is stumpy, the hind-limbs are distinctly longer than those of the Paca, and the two lateral toes have disappeared from the feet—a concomitant as it seems of the Agouti's greater powers of running. The three metatarsals are closely pressed together, and the foot is as it were on the way towards the highly-modified foot of the Jerboa. The fore-feet are, however, five-toed. The clavicle is rudimentary,^[40] whereas it is well developed in the Paca. The skull has not the peculiar modifications of that of the last-mentioned type. The sternum has seven pieces, and eight ribs reach it. A curious difference between this genus and the last is in the relative proportions of the regions of the intestine. The figures given by Tullberg for the two animals are—for Coelogenys, small intestine, 4800 mm.; caecum, 230 mm.; large intestine, 21,000 mm.;—for Dasyprocta aguti the same author gives: small intestine, 4200 mm.; caecum, 200 mm.; large intestine, 1000 mm. The Agouti, says Mr. Rodway,^[41] is as wily as the Fox. "If chased he will run along the shallows of a creek to hide his scent from the dogs, or swim over and back again several times for the same purpose. He never runs straight when pursued, but doubles, often hiding until a dog has passed, and then making off in a different direction. Like the fox he has been hunted for a very long period, and, like Reynard, has grown wiser with every generation."

Fig. 242.—Agouti. Dasyprocta aguti. × ¹⁄₁₀.

Fam. 5. Dinomyidae.—The genus Dinomys of Dr. Peters^[42] is a very little known and remarkable form from South America allied to the Capybara, the Chinchilla, and other South American Rodents. It is only known by a single example found wandering about a courtyard in a town of Peru. It is externally like, and of about the same size as the Paca, but has a hairy tail. The animal is four-toed and plantigrade; the ears are short, and the nostrils are -shaped. It is usually regarded as belonging to a separate family which will include but the one species, D. branickii.

Fam. 6. Chinchillidae.—This family, likewise South American, contains three genera,^[43] all of which agree in having long limbs, especially the hind-limbs, and a bushy and well-developed tail. The hair is exceedingly soft, hence the commercial value of "chinchilla."

The genus Chinchilla, containing but a single species, C. laniger, is a small and squirrel-like creature, living at considerable heights in the Andes. The eyes, as it is a nocturnal creature, are naturally large; and so also are the ears. The fore-feet have five toes, the hind-feet only four; they are furnished with feeble nails. The innermost toe of the hind-foot has a flat and nail-like claw. There are thirteen dorsal vertebrae, and the long tail has more than twenty. The clavicle is well developed, as in the other genera of this family. The large intestine of this animal is extraordinarily long; the proportions of the different regions of the gut are shown by the following measurements: small intestine, 820 mm.; caecum, 125 mm.; large intestine, 1340 mm. Such a disproportion between the large intestine and the small, to the advantage of the former, is a very strange fact in the anatomy of this Rodent.

The genus Lagidium (also called Lagotis), which includes "Cuvier's Chinchilla," is also a mountain dweller. There are several species of this genus, which differs from Chinchilla by the complete abortion of the thumb and of the great toe. The intestinal proportions are those of Chinchilla. The ears and tail are long. L. cuvieri measures 1½ feet in length.

Lagostomus, again, has but one species, L. trichodactylus. The animal has a tail about half the length of the body. The digits are reduced as compared with Chinchilla, there being but four on the fore- and three on the hind-feet. There are only twelve dorsal vertebrae, and seven ribs reach the sternum. In the skull a distinguishing mark from the last two genera is the separation of the infra-orbital foramen into two by a thin lamella of bone. The large intestine is between one-half and one-third the length of the small intestine, and thus differs much from that of Chinchilla.

Fig. 243.—Vizcacha. Lagostomus trichodactylus. × ¹⁄₁₀.

The Vizcacha lives in societies of twenty to thirty members,^[44] in a "village" ("Vizcachera"), a dozen or so of burrows, which intercommunicate. They lie at home during the day and come out in the evening. Their burrows, like those of the Prairie Marmot, harbour other creatures, which apparently live on amicable terms with the Vizcachas; such are the burrowing owl, a small swallow, and a Geositta. The Fox also affects these burrows, but then he ejects the rightful owner of the particular burrow which he selects. When the young Foxes are born the vixen hunts the Vizcachas for food. The Vizcacha has a most varied voice, producing "guttural, sighing, shrill, and deep tones," and Mr. Hudson doubts if there is "any other four-footed beast so loquacious or with a dialect so extensive." These animals are very friendly, and pay visits from village to village; they will attempt to rescue their friends if attacked by a Weasel or a Peccary, and to disinter those covered up in their burrows by man.

Fam. 7. Cercolabidae.—A number of the characters which differentiate this family from the Hystricidae or Ground Porcupines of the Old World are given under the description of the latter. The principal external characters are the prehensile tail, the admixture of spines with hairs, and the nature of the sole of the foot. In these points the New-World Cercolabidae differ from the Old-World Hystricidae. It is interesting to notice that in both families we have long-tailed and short-tailed forms. Cercolabes corresponds to Atherura or Trichys, and Erethizon to Hystrix.

The genus Erethizon, the "Urson" of Canada, has a short, stumpy tail. Its spines are almost hidden by enveloping hair. The fore-feet have four, the hind-feet five toes. The short tail of this creature is remarkable when we reflect upon its climbing habits. It appears, however, to be a weapon with which it strikes sideways at the enemy.

Fig. 244.—Brazilian Tree Porcupine. Sphingurus prehensilis. × ¹⁄₆.

Of the Neotropical genus Cercolabes (sometimes called Sphingurus, Synetheres, or Coendou) there are some eight or nine species, all found in Central and South America. The animal is arboreal, and has in correspondence with that habit a prehensile tail. The spines are not so stout as in the Ground Porcupines, and are often coloured yellowish or reddish. In correlation with its tree-frequenting habits the bones of Cercolabes show certain differences from those of the Ground Porcupines. The scapula is broader and rounder in front than is that of Hystrix; the phalanges of the thumb (which is rudimentary) are fused together as in the Canadian Erethizon; but those of the very small hallux are also fused, whereas in Erethizon, as in Hystrix, they are separate. In one species, C. insidiosus, Sir W. Flower states that there are as many as seventeen dorsal vertebrae and thirty-six caudals. The tail is thus very long. In C. villosus there are fifteen dorsals and twenty-seven caudals; eight ribs reach the sternum, which is composed of seven pieces, the sixth being very small. The clavicles are well developed. A curious fact about C. villosus is that the acetabular cavity is perforate (on both sides), or at least only closed by membrane. In many forms of Rodents the bone is very thin in this region. This fact perhaps lessens the significance of the perforation of the acetabulum of Echidna (see p. 109).

Of the allied genus Chaetomys, also Neotropical, there is but a single species, which inhabits Brazil. It has a nearly completely closed orbit, a feature which differentiates it from the last animal, and one which also shows it to be a more modified form. The spiny covering is less pronounced than in its allies.

Fam. 8. Hystricidae.—This family is characterised by the fact that all its members possess spines; but the tail, if at all long, is not prehensile, and the soles of the feet are smooth and not covered with rough tubercles, as in the Tree Porcupines of the next family, Erethizontidae. The clavicle is less developed than in the arboreal forms. In the organs of digestion there are points of a family difference between the two groups of spiny Rodents. The tongue has serrated scales arranged in transverse rows, which are directed backwards. A gall-bladder, though not always present, is sometimes found; it apparently never exists in the arboreal Porcupines and in Erethizon. The lungs show a great tendency to subdivision, which appears to be especially marked in the genus Atherura. The caecum seems also to be shorter in the Ground Porcupines. In Hystrix cristata the small intestine measures 15 feet 7 inches; the caecum, 8 inches; the large intestine, 4 feet 4 inches:—in Atherura africana the caecum measures 7½ inches; the large intestine, 1 foot 10 inches. The corresponding measurements of Synetheres villosus were: small intestine, 7 feet 3 inches; caecum, 1 foot 4 inches; large intestine, 2 feet 7 inches. In Erethizon the caecum is 2 feet 4 inches in length. These differences are too large and too constant in a number of presumedly allied forms to be overlooked.

Mr. Parsons has directed attention^[45] also to a number of muscular differences, such indeed as might be expected to occur between animals of such different habits.

The genus Hystrix embraces the better-known Porcupines. It is a genus of wide range, extending from the East Indies to Africa, and even occurring in Europe. There are several species, of which the common Hystrix cristata is the best known, and is the one which is to be found in Europe.

Fig. 245.—Common Porcupine. Hystrix cristata. × ¹⁄₁₀.

The spines of the common form and of the others are solid in the middle of the body, but on the tail they are expanded into hollow quills, which make much rattling. They are as a rule black and white, the middle of the spine being banded with black. A great crest of coarse long hairs on the head is responsible for the scientific name of the well-known form. Sometimes in this genus, as in the Tree Porcupines of Brazil, the spines are orange or yellow; but it is said that the colour is soon lost in this country. As a matter of fact it is the easiest thing in the world to wash out with ordinary tap-water much of the yellow colour of the spines of the South American Sphingurus. The same may be the case with the pigment of the Old-World Porcupines. There are fourteen to fifteen dorsal vertebrae and four or five lumbars. The tail varies in length, but is shorter than the long tail of the arboreal New-World forms. It seems impossible when mentioning the Porcupine to escape from some observations about its alleged habit of shooting its quills. For some reason or other Buffon has got the credit of inventing, or at least promulgating, this legend, which has even grown so in the telling that the quills are said to be capable of penetrating planks of wood. What Buffon said apropos of this matter is, "The marvellous commonly is pleasingly believed, and increases in proportion to the number of hands it passes through." It is of course the rattling of the spines and the occasional falling out of loose ones which has started the legend. They are, however, excellent weapons of offence, and the animal charges somewhat backwards to make the best use of them against the foe. The spines, however, are by no means an absolute protection, since, as Mr. Ridley informs us,^[46] Tigers will kill and eat these animals just as the Thylacine is apparently indifferent to the spiny armature of Echidna.

Of the Brush-tail Porcupine, Atherura^[47] there are at any rate two species, the West African A. africana and the Malayan A. fasciculata. It is interesting that the gap in the present distribution is partially filled by the discovery of fossil teeth near Madras. The genus does not differ widely in external appearance from Hystrix; it has, however, a rather longer tail; there are fewer large spines, and there is a tuft of them at the end of the tail, whence is derived the name of the genus. The frontal bones project a little distance between the nasals, a feature which does not seem to appear in the true Porcupines. There are fourteen dorsal vertebrae and five lumbars. The twenty-four caudal vertebrae of this Porcupine shows how much longer is its tail than that of Hystrix; for in the latter twelve is about the number.

A third genus of Old-World Porcupine is the singular Trichys.^[48] Of this there is but one species, T. lipura. It is a curious fact that out of three examples, all from Borneo, two were quite without a tail. But this appears to be merely a mutilation, though it is singular that the natives state it to be without a tail. One cannot help thinking of the way in which lizards sometimes shed their tails when pecked at. The tail of this genus is more than half the length of the body and head. Trichys has sixteen dorsal and six lumbar vertebrae. There is a tuft of quills at the end of the tail, which are thin and compressed, though truncate at the free extremity and hollow; they represent in a more rudimentary way the much stronger tuft at the end of the tail of other Porcupines. It is a curious fact that this and other Porcupines possess a mechanism for warning their foes precisely comparable to that of the rattlesnake. There are sixteen dorsal vertebrae.

Sub-Order 2. DUPLICIDENTATA.

The chief feature of this group is the existence of two pairs of incisor teeth in the upper jaw, of which the inner are very small and lie behind the outer. In the skull the infra-orbital foramen is small; the incisive foramina are very large. The tail is short or absent.

Fam. 1. Leporidae.—This family is distinguished from the Lagomyidae by the long ears, by the tail, which is present, though short, and by the longer limbs. There are six teeth belonging to the molar series in the upper jaw, and five of the same in the lower. The clavicle is imperfect.

The longest known genus of this family, Lepus, was, until the quite recent discovery of Romerolagus, the only genus. It is of universal range, excepting Australasia and Madagascar, and consists of about sixty species. These are the Hares and Rabbits, to the former being assigned the longer-limbed forms.

As every text-book of zoology contains a more or less elaborate account of the structure of the Common Rabbit, and as there is but little structural difference between the members of the genus, a short account of the generic peculiarities of Lepus will suffice here. The fore-feet are five-toed, the hind-limbs four-toed. The hairy integument enters the mouth cavity, and the inside of the cheeks have a hairy covering. The soles of the feet are, moreover, hairy. The maxillary bones are curiously sculptured.

The Common Rabbit, L. cuniculus, differs from the Common Hare in the comparatively shorter ears and legs. The ears have not, to so marked a degree, the black tips of those of the Hare. The animal, moreover, produces naked young, and lives in burrows of its own excavation. A difference in the structure of the caecum, which distinguishes the Rabbit from the Hare, has been

Fig. 246.—Lepus cuniculus. Skull. A, Lateral view; B, ventral view. ang.pro, Angular process of mandible; as, alisphenoid (external pterygoid process); aud.me, external auditory meatus; b.oc, basioccipital; b.sph, basisphenoid; cond, condyle; cor, coronoid process; fr, frontal; int.pa, inter-parietal; ju, jugal; lcr, lachrymal; max, maxilla; nas, nasal; opt.fo, optic foramen; o.sph, orbitosphenoid; pa, parietal; pal, palatine; pal.max, palatine plate of maxilla; pal.p.max, palatine process of premaxilla; par.oc, paroccipital process; peri, periotic; p.max, premaxilla; pt, pterygoid; p.t.sq, post-tympanic process of squamosal; s.oc, supraoccipital; sq, squamosal; ty.bul, tympanic bulla; vo, vomer; zyg.max, zygomatic process of maxilla. (From Parker and Haswell's Zoology.)

pointed out by Professor W. N. Parker.^[49] These differences have led some to approve of its separation from the Hares into a genus Oryctolagus. This animal is believed to be an introduced species, and to have been brought by man into these islands. Its original home is the Spanish Peninsula, the south of France, Algiers, and some of the Mediterranean islands. Mr. Lydekker thinks that the only other species of Lepus which can be considered to be a "Rabbit" is the Asiatic L. hispidus.

Of Hares there are two species in this country. The Common Hare, L. europaeus (the name L. timidus seems to be really applicable to another species to be referred to presently), extends all over Europe excepting the extreme north of Russia and Scandinavia. It is not known in Ireland, and, curiously enough, attempts to acclimatise this animal in that island have failed—a state of affairs which contrasts with the fatal ease with which the Rabbit has been introduced into Australia. Ireland has, however, the Variable Hare, L. timidus (also called L. variabilis), a species which is common in other parts of Europe, and which extends as far east as Japan. This species differs from its ally by the fact that it often turns white in winter with the exception of the black tips to the ears. In Ireland this change does not always occur; but Mr. Barrett-Hamilton has commented upon the fact that Hares of this species do change on Irish mountains. It appears that in this animal the change from the winter to the summer dress is accomplished by the actual casting off of the white hairs and their replacement by a fresh growth of "blue" hairs. A similar change occurs in the American L. americanus.

Dr. Forsyth Major has noted the fact that the various species of Hares can be distinguished by the condition of the furrows upon the upper incisors. Thus two African species, L. crawshayi and L. whytei, are to be separated by the fact that in the former the incisors are quite flat, whereas in L. whytei the groove is more prominent and there is a second shallow furrow.

The genus Romerolagus^[50] is quite a recent discovery. It occurs on the slopes of Popocatepetl in Mexico; it has the general aspect of the last genus, and is spoken of as a "Rabbit." It inhabits runs in the long grass which clothes the sides of the mountain. Externally it is something like the Pikas, since it has no tail visible. The ears, too, are short, and the hind-legs comparatively short. The skull is very like that of the Rabbit; but in other osteological details it is aberrant. Thus the clavicle is quite complete, and only six ribs articulate with the sternum, instead of the seven that we find in the Rabbit.

Fam. 2. Lagomyidae.—The animals of this family are smaller than the Hares and Rabbits; they have short Vole-like ears and no external tail. The limbs also appear to be shorter. As there is but a single genus, the characters of the family may be described in connexion with those of the genus, which is known as Lagomys (apparently more correctly Ochotona). Of this genus there are about sixteen species, which are mainly Asiatic; one species extends its range into Eastern Europe, and three are North American.

The skull has not the supra-orbital grooves of the Rabbits, and has a well-marked backward process of the zygomatic arch. There are eighteen dorsal vertebrae. The molars and premolars are five.

The vernacular names of "Pika" and "Piping Hares" have been applied to the members of this genus, the latter on account of their peculiar call. They live among rocks in companies and they burrow. They are usually found at considerable altitudes: thus L. roylei, the "Himalayan Mouse Hare," is found at elevations as high as 16,000 feet; while L. ladacensis gets even higher, 19,000 feet having been recorded. With the habits of a Marmot, so far as concerns living in burrows and at great altitudes, the animals of this genus, with their squat form and short ears, are not unlike those animals. In the past this genus occurred more generally over Europe. Species from Miocene beds have been met with in England, France, Germany, and Italy.

Fossil Rodents.—Quite a large number of existing genera of Rodents are known from even the earlier strata of the Tertiary period. The Squirrels (and even the genus Sciurus itself) occur in the Upper Eocene. So, too, do the genera Myoxus, and (in South America) Lagostomus. Spermophilus, Acomys, Hystrix, Lagomys, Lepus, Hesperomys are known from Miocene rocks. Rhizomys, Castor, Cricetus, Mus, Microtus, and some others appear to have originated so far as we know in the Pliocene, while a still larger series of existing genera are Pleistocene. It is interesting to note that some of the extinct genera were much larger than recent forms. At present, Hydrochoerus is the biggest Rodent; but the genus Megamys from the Pampas formation of Argentina was "nearly as large as an ox." The wider range of genera in the past is illustrated by Hystrix, which, now an Old-World form, is represented by remains in the Miocene and Pliocene of America.

It is a significant fact that of living genera Sciurus is the oldest; for it has been pointed out that in a number of features the Squirrels are among the most primitive of Rodents. The zygomatic arch is slender, and has thus not acquired the specialisation that is to be found in that part of the skull in other Rodents; moreover, the "jugal bone is not supported by any process from the maxilla exactly as in the primitive Ungulata." The feet, too, are unspecialised, though that is the case with many other genera. It may also be pointed out that the teeth bear not a little likeness to those of Ornithorhynchus in their multituberculate character.

Some few fossil forms have already been dealt with in the preceding pages.

The two genera Castoroides and Amblyrhiza, from the Pleistocene of North America and the West Indies, are usually regarded as forming a family. The skull of the former genus indicates an animal of the size of a Bear. It is compared to that of Castor, but it has a wide infra-orbital foramen. The teeth are four in each jaw, and are formed of three to five lamellae; the incisors of this animal are powerful but short. Amblyrhiza, on the other hand, has long incisors which are longitudinally grooved anteriorly. It has a free fibula. This latter as well as other characters have led Tullberg to remove it from association with Castoroides.

Order X. TILLODONTIA.

This group of Eocene mammals is to be defined by a number of characters, of which the more important are the following:—The incisors are enlarged, grow from persistent pulps, and are coated with enamel upon the outer surface only; they are those of the second pair only, the first and third having disappeared or become small. The canines are reduced in the later forms.

Fig. 247.—Tillotherium fodiens. Left lateral view of skull. (From Flower, after Marsh.)

These animals have been regarded as ancestral Rodents, to which the tooth characters just mentioned clearly show likenesses. The earliest known form is Esthonyx. This genus shows such primitive characters, compared with its later representatives, as the existence of all three pairs of incisors in the upper jaw, but only two in the lower jaw. The enlarged incisors of both jaws do not seem to have grown from persistent pulps.

Anchippodus, a later form, still preserves the upper pair of first incisors in a vestigial form; the strong second incisors grew from persistent pulps. The most recent genus, Tillotherium, shows the characteristics of the group at their height. The strong Rodent-like, chisel-shaped incisors, which are reinforced by a small additional pair in the upper jaws only, are persistent. The grinding teeth are of the tritubercular pattern; there are three of each kind in the upper jaw, but in the lower jaw only two premolars on each side. This is at any rate the case with some, while others have three. The canine, though present in both jaws, is insignificant. As in many ancient types, there is an entepicondylar foramen in the humerus. The feet were five-toed, and bore sharp, laterally-compressed claws. The skull has been compared in general aspect to that of a Bear.

---

1. See especially Tullberg, "Ueber das System der Nagethiere," Act. Ak. Upsala, 1899; and Alston, Proc. Zool. Soc. 1875, p. 61; and for nomenclature, Thomas, Proc. Zool. Soc. 1896, p. 1012; and Palmer, Proc. Biol. Soc. Washington; xi. 1897, p. 241.
2. Proc. Zool. Soc. 1884, p. 252.
3. Phil. Trans. 1850, pt. ii. p. 529.
4. Seen, however, in Chaetomys.
5. See Beddard, Proc. Zool. Soc. 1892, p. 596, and Gervais, Journ. Zool. i. 1872, p. 450.
6. "Observations sur le genre Anomalurus," Nouv. Arch. Mus. (2), vi. 1883, p. 277.
7. "On the Habits of the Flying Squirrels of the genus Anomalurus," Proc. Zool. Soc. 1894, p. 243.
8. W. E. de Winton, "On a New Genus and Species of Rodents," etc., Proc. Zool. Soc. 1898, p. 450. Apparently just at the time of the publication of this paper Matschie described the same animal as Zenkerella.
9. Proc. Zool. Soc. 1893, p. 179.
10. Flower and Lydekker.
11. Thomas, J. Asiat. Soc. Bengal, lvii. 1888, p. 256.
12. E. T. Newton, Trans. Zool. Soc. xiii. 1892, p. 165.
13. Proc. Zool. Soc. 1896, p. 1016.
14. Reuvens, "Die Myoxidae oder Schläfer," Leyden, 1890, allows but one genus, Myoxus, the other genera adopted here being termed subgenera.
15. To which a sixth, the "Yellow-necked Mouse," Mus flavicollis, may perhaps be added.
16. For anatomy see Windle, Proc. Zool. Soc. 1887, p. 53.
17. Proc. Zool. Soc. 1889, p. 247.
18. Trans. Zool. Soc. xiv. 1898, p. 377.
19. Nouv. Arch. Mus. iii. 1867, p. 81.
20. Popular Natural History of Animals, London, 1898.
21. Proc. Zool. Soc. 1863, p. 95.
22. See O. Thomas, "On some Mammals from Central Peru," Proc. Zool. Soc. 1893, p. 333.
23. "Notes on the Rodent genus Heterocephalus," Proc. Zool. Soc. 1885, p. 845.
24. Proc. Zool. Soc. 1885, p. 611.
25. Proc. Zool. Soc. 1890, p. 610.
26. Parsons, Proc. Zool. Soc. 1898, p. 858.
27. Very probably this form should be rather, as it is by Thomas, referred to the neighbourhood of Pectinator, which would clear up the geographical anomaly.
28. Notes Leyd. Mus. 1891, p. 105.
29. Günther, Proc. Zool. Soc. 1879, p. 144.
30. Proc. Zool. Soc. 1873, p. 786.
31. Loc. cit. (on p. 458), p. 123.
32. Proc. Zool. Soc. 1892, p. 520.
33. See Dobson, Proc. Zool. Soc. 1884, p. 233.
34. Peters, Trans. Zool. Soc. vii. 1871, p. 397.
35. "Field Notes on the Mammals of Uruguay," Proc. Zool. Soc. 1894, p. 297.
36. Beddard, Proc. Zool. Soc. 1891, p. 236.
37. These are stated by Tullberg to be absent. I have found them, but they are very small bones, not more than half an inch long.
38. There is a faint development of these ridges, but behind the palatine foramina in Dasyprocta aguti.
39. Proc. Zool. Soc. 1885, p. 161.
40. Or absent?
41. In the Guiana Forest, London, 1894.
42. MB. Ak. Berlin, 1873, p. 551.
43. An account of the three genera is to be found in Trans. Zool. Soc. i, 1833, p. 35, by Mr. E. T. Bennett.
44. Hudson, "On the Habits of the Vizcacha," Proc. Zool. Soc. 1872, p. 822.
45. Proc. Zool. Soc. 1894, pp. 251, 680.
46. Nat. Science, vi. 1895, p. 94.
47. See Parsons. Proc. Zool. Soc. 1894, p. 675.
48. Günther, Proc. Zool. Soc. 1876, p. 739, and 1889, p. 75; and Cederblom, Zool. Jahrb. Syst. Abth. xi. 1897-98, p. 497.
49. Proc. Zool. Soc. 1881, p. 624.
50. Proc. Biol. Soc. Washington, x. 1896, p. 169.